An analytical and numerical study has been carried out with the view on the understanding of the physical mechanisms of the mixing process in a gas turbine burner. To this end, three methods at various levels of approximation have been used: At the simplest level an analytical model of the burner flow and the mixing process has been developed. It is demonstrated how this approach can be used to understand basic issues of the fuel-air mixing and how it can be applied as a design tool which guides the optimisation of a fuel injector device. At an intermediate level of approximation, steady-state CFD simulations, based on the k–ε- and RSM-turbulence models are used to describe the mixing process. All steady simulations fail to either predict the recirculation zone or the turbulence level correctly, and can therefore not be expected to capture the mixing correctly. At the most involved level of modelling time-accurate CFD based on unsteady RSM and LES-turbulence models are performed. The simulations show good agreement with experiments (and in the case of LES excellent agreement) for both, velocity and turbulence fields. Mixing predictions close to the fuel injectors suffer from a simplification used in the numerical setup, but the mixing field is predicted very well towards the exit of the burner. The contribution of the asymmetric coherent flow structure (which is associated with the internal recirculation zone) to the mixing process is quantified through a triple decomposition technique.

This content is only available via PDF.
You do not currently have access to this content.